基于PSO的板形板厚小波神经网络解耦PID控制

针对板形控制和板厚控制是相互耦合的综合系统,提出了一种新的解耦PID控制算法·首先用小波神经网络构造α阶时延逆系统,对综合系统进行输入输出解耦;然后对解耦后的独立的单变量系统采用PID控制·这种解耦方法无论是从理论分析还是仿真验证,均证明是可以实现完全解耦的·考虑到被控对象是一个带有时滞的非线性系统,提出采用PSO优化算法对PID参数进行自适应调整·仿真结果表明所用方法简单有效,并具有良好的跟随性能和抗干扰能力;其控制效果优于传统的解耦PID控制·

PSO-Based Decoupling PID Control Using Wavelet Neural Network for Strip Flatness/Gauge

Automatic flatness control (AFC) and automatic gauge control (AGC) are interacted and coupled with each other. A novel decoupling PID control method for AFC-AGC is presented. The α-order time-delay inverse systems based on wavelet neural networks (WNN) are built and used as compensators for input/output decoupling of AFC and AGC. Then, PID controller is adopted to control the SISO systems. Theoretical analysis and numerical simulations show that the decoupling method proposed is able to decouple completely. The parameters of PID are adaptively adjustable for non-linear system with time-delay if using particle swarm optimization (PSO) algorithm. Simulation results show that the control system is simple and effective and has good performance of adaptively tracking target and resistance to disturbances. It is superior to conventional decoupling PID control to improve the accuracies of strip flatness and gauge.